The diffuse, infiltrative nature of high-grade gliomas is a major obstacle to curing these tumors. In order to make a significant impact on survival, new treatment strategies must specifically target these extremely invasive tumor cells. Human neural stem cells (NSCs) hold great promise for glioma therapy because of their inherent ability to target tumor cells throughout the brain. By harnessing their tumor-tropism and genetically modifying them to express a therapeutic transgene, NSCs can act as delivery vehicles for targeted anti-cancer therapies. NSC-mediated treatment approaches can potentially increase tumor-selectivity, decrease toxicities, and achieve therapeutic indices sufficient to eradicate invasive and residual tumor cells. We propose the use of a well-characterized, clonal, allogeneic NSC line (HB1.F3) that has been retrovirally-transduced to stably express the enzyme, cytosine deaminase (CD), which converts the oral prodrug 5- fluorocytosine (5-FC) to the chemotherapeutic agent 5-fluorouracil (5-FU). Preliminary data in normal and orthotopic glioma murine models indicate that this cell line is safe, non-tumorigenic, non-immunogenic, and therapeutically active. In Specific Aim 1 of this research project, a pilot feasibility study in recurrent high-grade glioma patients will be performed to determine the safety of intracerebral administration of HB1.F3.CD NSCs in combination with 5-FC. Three dose levels of NSCs will be tested. Specific Aim 2 will demonstrate proof-of-concept by assessing the extent to which the CD-expressing NSCs convert 5-FC to 5-FU at sites of tumor. Intracerebral levels of 5-FC and 5-FU will be measured by micro dialysis, and we will characterize the relationship between intracerebral and systemic concentrations of 5-FC and 5-FU with increasing NSC dose level. 19F MRS will also be used to non-invasively document the presence of 5-FU in the brain during 5-FC treatment. PUBLIC HEALTH RELEVANCE: Human fetal neural stem cells (NSCs) are inherently tumor-tropic. When modified to express a therapeutic transgene, NSCs have the potential to be used as delivery vehicles for anti-cancer therapies. We propose the clinical use of a well-characterized, clonal, allogeneic NSC line, HB1.F3, which has been modified to express cytosine deaminase (CD). CD converts the inactive prodrug 5-fluorocytosine (5-FC) to the active chemotherapeutic 5-fluorouracil (5-FU). Preliminary biodistribution studies indicate this cell line is safe, non-tumorigenic and non-immunogenic. HB1.F3.CD NSCs have demonstrated efficacy in animal glioma models. Specific Aim 1 of this research proposal is to determine the safety of intracerebral administration of HB1.F3.CD NSCs in combination with oral 5-FC, in patients with recurrent high-grade gliomas. Specific Aim 2 will assess the extent to which the HB1.F3.CD NSCs convert 5-FC to 5-FU at sites of tumor.